What's Important: Time-Independent Schrödinger Equation

PHYS 385 Lecture 9 - Time-independent Schrödinger equation 9-1
Lecture 9 - Time-independent Schrödinger equation
What's important:
• time-independent Schrödinger equation
• eigenfunctions and eigenvalues
• particle in a 1D box
Text: Gasiorowicz, Chap. 4
Time-independent Schrödinger equation
Let's apply the ideas of eigenfunctions to the time-dependent Schrödinger equation to

extract the time-independent Schrödinger equation. In this, and the next several
lectures, we continue to work in one dimension. Recall the explicit representation of the
Schrödinger equation:
(x,t ) h2 2 (x,t)
ih =− +V (x ) (x,t ), (1)
t 2m x2
where we have made the potential energy a function only of position. If an equation can
be segregated into parts which depend on only one variable, then a fruitful approach to
solving it is the so-called separation of variables approach. As implemented here, that
means we should try to write the wavefunction (x,t) as a product of two functions, one
containing all the position dependence and one containing all the time dependence:
(x,t) = T(t)•u(x). (2)
Aside: not everyone uses this notation; frequently one sees (x,t) for the complete
wavefunction, and (x) for the position dependent part.
Substituting (2) into (1), and shuffling some functions to show what derivatives need
attention
dT (t) h2 d 2u(x)
ihu (x) =− T (t) +T (t )V (x)u(x).
dt 2m dx 2
Note how ∂ has become d. Dividing by T(t)•u(x) yields

dT(t) /dt −(h2 / 2m)(d 2u(x )/ dx 2 ) +V (x)u(x)
ih = (3)
T(t ) u(x)
Now, the time dependence is completely isolated on the left hand side, and the position
dependence is on the right hand side. Since the LHS cannot depend on x, and the RHS
cannot depend on t, then NEITHER side can depend on x or t. That is, both sides must
be equal to a constant, which we define with some foreshadowing as E. Treating the
LHS first, we can now write
©2003 by David Boal, Simon Fraser University. All rights reserved; further copying or resale is strictly prohibited.
dT(t)
ih = E T (t ) (4)
dt
This has the obvious solution

T (t) = Ce− iEt / h (5)
The RHS cannot be explicitly solved until we know the functional form of V(x).
However, it can be explicitly written out as
h2 d 2u(x )
− 2 +V (x)u (x) = E u(x) (6)
2m dx
Eq. (6) is the time-independent Schrödinger equation and it is now obvious why we
chose to call the constant E, the energy. Eq. (6) is an eigenvalue equation.
Eigenfunctions and eigenvalues
Eq. (4) is an example of an eigenvalue equation. It states that if we operate on the

function T(t) with the operator i h d /dt, we regain the function multiplied by a constant,
namely the energy E. Here, E is the eigenvalue and T(t) is the eigenfunction.
Gasiorowicz reviews a number of properties of eigenvalues etc. which the student is
encouraged to review. The eigenvalue equation selects both the eigenvalue and the
eigenfunction from all of the values and functions available.
Hopefully, most of this material has been seen in MATH 232, and there is no need to
establish them again. An analogy emphasized by Gasiorowiz is the rotation of a vector
in three dimensions. Calling the vector A and the rotation matrix R( ) for a rotation by
around the z-axis, in general the operation
R( )A
carries A to a new vector B having the same length as A but pointing in a different
direction. An eigenvalue equation like
R( )A = cA
for ≠ 0, π/2, π... would lead to

c = +1
A pointing along the rotation axis.
In other words, both c and A are selected by the equation. If we choose the special
values for (like = π/2), a more general set of vectors A might be selected.
Particle in a 1D box
We referred to this problem in Lec. 1, in describing a particle travelling in one dimension

between hard walls; the standing wave condition on the de Broglie wavelength lead to a
set of solutions satisfying
L = n /2 = nh / 2p n = 1, 2, 3...
or
p = nπ h /L. (7)
We now revisit this problem with the time-independent Schrödinger equation (6)
h2 d 2u(x )
− +V (x)u (x) = E u(x)
2m dx 2
for the potential

V=0 0<x<a
V = +∞ x < 0 and x > a.
V(x)
Eq. (6) tells us that the wavefunction u(x) must vanish at x = 0 and a where the potential
becomes infinite. In the intervening range of x, the wavefunction must satisfy
d 2u /dx 2 = - (2mE / h 2) u. (8)
Now, if we define
k 2 = 2mE / h 2 (9)
then Eq. (8) has the particularly simple form

d 2u /dx 2 = -k 2u. (10)
This equation has the usual sine and cosine functions as a solution. The cosine set
cos(kx), where the argument has been made dimensionless as appropriate, does not
satisfy the boundary condition u(0) = 0. Hence, only the sine functions are admissible:
u(x) = A sin(kx)
where A is a normalization constant. The second boundary condition u(a) = 0 places a

constraint on k via
ka = nπ n = 1, 2, 3, ... (11)
and the normalization condition to make u(x) a probability density

∫oa u*(x) u(x) dx = 1
leads to
1 = A 2 ∫oa sin2(nπx /a) dx
= A 2 (a / nπ) ∫onπ sin2 d
= A 2 (a / nπ) n (π/2) (π/2 from the usual ∫oπ sin2 = π/2)
= A 2 a/2.
Thus
A = (2/a)1/2. (12)
Gasiorowicz further establishes that the wavefunctions are orthogonal (see p. 59)
∫oπ un*(x) um(x) dx = 0 n≠m
by evaluating the integral with the trig identity

sina sinb = cos(a-b) - cos(a+b).
In summary, the solutions

un = (2/a)1/2 sin(nπx /a)
form an orthonormal set of functions - they are both normalized and mutually
orthogonal:
∫oa un*(x) um(x) dx = δnm. (13)
Note that in general, even if a set of solutions contains some functions with the same
energy (called degenerate solutions), they are nevertheless orthogonal.
What about the energy and momentum of these solutions? Now that k is known from
Eq. (11), the energy in Eq. (9) becomes
En = h 2k 2 / 2m
= π2 h 2 n 2 / 2ma 2, (14)
which is the same as we found from the de Broglie condition. This still doesn't tell us p2,
which must be found from the expectation
<p 2> = ∫oa u*(x) p 2 u(x) dx = − h 2 (2/a) ∫oa sin(kx) (d /dx)2 sin(kx) dx
= − h 2 (2/a) (-k 2) ∫oa sin(kx) • sin(kx) dx
= ( h k)2.
Again, we regain the de Broglie expression.
A couple of miscellaneous issues:
1. The expression for the kinetic energy can be rearranged as follows:

K = <p 2> / 2m = -( h 2 / 2m) ∫ u*(x) (d /dx)2 u(x) dx
h2  d  du  du * du 
2m ∫ dx  dx  dx dx 
=− u* − dx
As usual, both u and its derivative vanish at infinity, so the first term, after performing
the ∫ (d /dx) dx must vanish. This leaves
K = ( h 2 / 2m) ∫ |du /dx|2 dx, (15)
showing that the kinetic energy is proportional to the complex square of the curvature of
the wavefunction.
2. We showed in an earlier lecture how a periodic function could be expressed as a

Fourier series - a sum over a set of functions sin(kx) just like the u's here. In other
words, a general wavefunction for this problem could have the form
(x) = Σn An un(x). (16)
Gasiorowicz shows that |An|2 is the probability that the system will be found to have an
energy En during a measurement.

What's Important: Time-Independent Schrödinger Equation

Uploaded by

Copyright:

Available Formats

What's Important: Time-Independent Schrödinger Equation

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

What's Important: Time-Independent Schrödinger Equation

Uploaded by

Copyright:

Available Formats

PHYS 385 Lecture 9 - Time-independent Schrödinger equation 9-1

Lecture 9 - Time-independent Schrödinger equation

Time-independent Schrödinger equation

Let's apply the ideas of eigenfunctions to the time-dependent Schrödinger equation to

Note how ∂ has become d. Dividing by T(t)•u(x) yields

This has the obvious solution

Eigenfunctions and eigenvalues

Eq. (4) is an example of an eigenvalue equation. It states that if we operate on the

for ≠ 0, π/2, π... would lead to

We referred to this problem in Lec. 1, in describing a particle travelling in one dimension

for the potential

then Eq. (8) has the particularly simple form

where A is a normalization constant. The second boundary condition u(a) = 0 places a

and the normalization condition to make u(x) a probability density

by evaluating the integral with the trig identity

In summary, the solutions

Again, we regain the de Broglie expression.

A couple of miscellaneous issues:

1. The expression for the kinetic energy can be rearranged as follows:

2. We showed in an earlier lecture how a periodic function could be expressed as a

You might also like